CHLORIDE-CATALYZED IRON(II)-IROK(III)ELECTRON EXCHANGE
Nov. 5, 1964 [CONTRIBUTION FROM
THE
CHEMISTRY DEPARTMENT, BROOKHAVEX NATIONAL LABORATORY, UPTON,
4591
XEW \'ORK]
The Inner-Sphere Activated Complex for the Electron Exchange of Iron(I1) and the Monochloro Complex of Iron(II1) BY R . J. CAMPION, T. J. COKOCCHIOLI, A N D N . SUTIN RECEIVED J U N E 23, 1964 The kinetics of the iron(I1)-catalyzeddissociation of FeCIZChave been studied by the use of a flow technique. The rate constant for the dissociation is given by k d = kl kn/(H-) k3(FeZL), where kl = 1.1 f 0.2 set.-', kr = 3 . 4 f 0.6 N sec.-l, and ka = 12.1 f 1.6 M-l sec.-l a t 25,0° and ionic strength 3.0 M . The kinetics of the iron! 11)-iron( 111) electron exchange in the presence of chloride ions were studied using a radioactive tracer method. The rate constant for the electron exchange between FeCl*- and F e z + was found to be 57.6 2 k-l set.-' a t 25.0' and ionic strength 3.0 M . The results suggest that the electron exchange between FeCI2' and F e z + proceeds mainly by a chloride-bridged inner-sphere activated complex.
+
+
+
Introduction T h e iron(I1)-iron(II1) electron exchange is catalyzed by a number of anions.2-8 This catalysis is generally ascribed to reactions between iron(I1) and complexes of iron(II1) with the added anions. The exchange reactions may proceed via electron-transfer or atom-transfer mechanisms, and indirect evidence for both types of mechanisms has been advanced.8-10 In the case of the exchange between iron(I1) and the monochloro complex of iron(III), the electron-transfer and atom-transfer mechanisms may be represented as
FeClZ+
+ F e z + / FeC1-
Fe3+ 1 FeClZ++ F e z +
(1)
+
(2)
where reactions 1 and 2 involve transfer of an electron and a chlorine atom, respectively. In the course of recent studies of the oxidation of F e 2 + by Co3+ in the presence of chloride ions," i t was observed that the rate of dissociation of the FeC12+ produced in the reaction depended upon the ferrous ion concentrations of the solutions. Because of the implication that this observation has for the mechanism of the chloride-catalyzed iron(I1)-iron(II1) exchange, we have made a more detailed study of the effect of F e z + on the rate of dissociation of FeCl*+. Comparisons of the rates of the iron(I1)-catalyzed dissociation of FeC12+ and the chloride-catalyzed iron(11)-iron(II1) electron exchange suggest strongly t h a t the electron exchange between F e 2 + and FeC13+ proceeds mainly via a chloride-bridged inner-sphere activated complex. Experimental T h e iron(II1) perchlorate was obtained from the G . Frederick Smith Chemical Co. and was purified by recrystallization from perchloric acid. Solutions of iron( 11) perchlorate were prepared by electroreduction of perchloric acid solutions of iron( 111). (1) Kesearch performed under t h e auspices of the U. S. Atomic Energy Commission. (2) J . Silverman and R . W . Dodson, J . P h y s . C h e m . , 66, 846 (1952) ( 3 ) J , Hudis a n d A . C. W a h l , J . A m . Chem. Soc., 76, 4153 (1953) (4) R . A H o r n e , J . P h y s . C h e m . , 64, 1512 (1960). (.>I G. S . Laurence, T ~ Q Y FZaSv.a d a y Soc., 63, 1326 (1957). (6) D . B u n n , F . S. Dainton, and S. D u c k w o r t h , ibid., 6 7 , 1131 (1961). (7) N. S u t i n , J , K. Kowley, and K. W. Dodson, J . P h y s . C h e m . , 66, 1248 (1961). (8) S .S u t i n , A n n . Reu. S u c l . S c i . , l a , 285 (1962). (9) H . T a u b e , Adoan. Iizorg. Chem Radiochem., 1, 1 (1959) (10) J. Halpern. Quaut. Rev. [ L o n d o n ) , 16, 207 (1961). (11) T. J . Conocchioli, G. S a n c o l l a s , and S .S u t i n , J . A m . Chem SOL.,8 6 , 1453 (1964)
The labeled iron(II1) perchlorate used in the exchange studies was prepared according t o the procedure of Hudis and Wahl.3 A solution of iron(II1) in 8 ,k! hydrochloric acid containing the Fejg tracer was extracted into isopropyl ether and back-extracted into water. Concentrated perchloric acid was added t o the aqueous phase, which was heated until fumes were evolved. The fuming procedure was repeated until a negative test for chloride was obtained. The perchloric acid (70co) and t h e hydrochloric acid (37.8%) were obtained from the J . T . Baker Chemical Co. Solutions of niagnesiuni, cobalt( 11), nickel(II), and manganese(11) perchlorate were prepared by neutralizing magnesium oxide (Baker analyzed reagent), cobalt carbonate (Baker and Xdamson), nickel carbonate, and manganese carbonate (Baker analyzed reagent), respectively, with perchloric acid. The kinetics of the dissociation of FeC12+ were studied using the apparatus and techniques described previously . l z - I 4 Solutions containing iron( 111), chloride, and perchloric acid were mixed with perchloric acid solutions containing either magnesium, cobalt(II), nickel( 11), manganese(II), or iron( 11) perchlorate and the disappearance of FeC12' followed a t 336 tnp. Some runs were also made in which the iron(II1) and chloride were put in different solutions, and the formation of FeC12- was followed. The same rate constants were obtained as in the dissociation measurements. The iron(I1)-iron(II1) electron exchange was followed using a modification of the procedure described by Silverman and Dodson.2 The exchange reaction was quenched by adding aliquots t o a solution containing 2,2'-dipyridine and potassium acetate. The dipyridine complexed the ferrous ion, and in addition potassium perchlorate precipitated. This precipitation reduced the perchlorate concentration t o a level sufficiently- low for the sparingly soluble perchlorate salt of ferrous dipyridine t o remain in solution. Iron(II1) carrier was added and ferric hydroxide precipitated with ammonia. The precipitate was filtered off and the specific activity of the ferrous dipyridine complex in the filtrate determined by measuring the Fej9 activity with a scintillation counter and determining the concentration of the complex spectrophotometrically a t 520 tnp. The kinetics of the dissociation of FeC12- and the iron( 11)ironCII1) electron exchange were studied a t 23.0" and a n ionic strength of 3.0 M.
Results and Discussion The Iron(I1)-Catalyzed Dissociation of FeC12+.-The dissociation of FeC12+ was studied in the presence of Co2+,X g 2 + ,Xi2-, A h 2 + , and F e z +ions. T h e chloride concentration was varied from 5.1 X 10Y4 to 5.1 X 10Y3 .If and the ironiII1) concentration from 5 . 2 X 10F4 t o 5.2 X 1W3 M. Since the degree of complex formation is relatively small, the concentrations of uncomplexed iron(II1) and chloride remained essentially constant during a particular run. The rate of dissociation was found to be first order with respect to FeCl*+ and independent of the iron(II1) and chloride concentrations of the solutions. Values of the rate constants for the dissociation, calculated from the halftimes for the reactions, are presented in Table I. (12) G. Dulz a n d S S u t i n . i n o v g . Chem.. a, 917 (1963: 113) D. Seewald and N . S u t i n . ibid , 2 , 6.13 (1963). (14) G. H. Nancullas a n d S .S u t i n , ibid , 3, :360 (1964)
R. J. CAMPION, T. J. CONOCCHIOLI, A N D N . SUTIN
4592 --+ r+
4.0
I
Id
Vol. 86
r
I
I
i 7 0
2
20
L,l
1.00.0
Fig. 1.-Plot
of
kd
'
'
0 30
' 0I50 I H C t 0,r' , M - '
'
'
0 40
1
0 60
'
01 70
/
3.04 A
os. (HC104)-' a t 25.0' and ionic strength 3.0 d l .
TABLE I
'-
o _ 0 10
It will be seen t h a t Co2f, Mg2+,Ni2+,and h I n 2 + all have the same relatively small effect on the rate of dissociation of FeC12+, whereas F e z + has a much larger effect. The dissociation in the absence of F e z + will be considered first. The increase in the rate of dissociation with increasing Co2+,hIg2+. Ni2+,and SIn2' concentrations can be interpreted in terms of the decreas-
I,. 3.0
0 20
(
Fig. 2.--Plots
0 30
040
0 50
O
Fe2'I. M
of k,i and [ k d - k l - k % / ( H + ) us. ] ( F e z + )a t 25.0" and ionic strength 3.0 -\I.
from the intercept and slope of Fig. 1 are 1.1 + 0.2 set.-' and 3 4 f 0.0 .\I set.-', respectively, a t 25.0" and ionic strength 3.i) .\I. h similar two-term rate law has previously been found by Connick and Coppel'j for the dissociation of FeC12 in perchloric acid-sodium perchlorate media a t an ionic strength of 1.0 M. They found that kl and k2 are 2 . 3 + 0.2 set.-' and 4.5 + 0 . 5 III set.-', respectively, a t 25.0" and ionic strength 1.0 +
THEEFFECT OF DIVALEST CATIOXS O N THE RATEOF DISSOCIATION OF FeCI2+ AT 25 0 ' A N D IONICSTRENGTH 30 W (HC101),
(M2+),"
M
.w
2.85 2.70 2.40 2 10 1.80 1 50
0.05 0 10 0 20 0.30 0.40 0.50
7
Mgz
-
p
cos+
+
2 39 2.57 2.66 2.73 2.92 3 44
2.35 2.48 2 51
k c i , sec. ----I;L-i2+
>fn'i
2 37 2.44 2 65
2.45 2.18 2 73
3 18 3.57
3.07 3.94
..
3 08 3.4i
Fez+
3 15 3 63 5 06 6 50 7 53 9 12
-If. The rate of disappearance of FeC12+in the presence of F e z +is given by
I[(FeCl*+)- (FeCl2+),,1
(M") is the concentration of the divalent cation during the reaction
(8)
with ing perchloric acid concentrations of the solutions since the measurements were made a t a constant ionic strength. Thus it is evident from Fig. 1 that a linear relationship is obtained when the average value of k d for a given concentration of the above cations is plotted against (HClO4)-'. The kinetic data in the absence of Fez+ are thus consistent with the rate law d(FeC12+) - [k, dt
+ $&)][(FeC12+)
where ka is the rate constant for the reaction
- (FeC12+),,]
(3)
with
where t l , , is the half-time for the dissociation of FeCl", and k 1 and kz are the rate constants for the reactions (H20),FeCI2-
+ H20
ki
+
F e ( H 2 0 ) 6 3 T C1-
(5)
+
(A)
k i
and (H20),FeC12+
+
(H10)4FeOHC1C H + K A k2'
(HJO)4FeOHC1+ HiO
where k2
=
K&'
+
( H 2 0 ) D F e O H 2 + C1-
(71
The values of kl and kl calculated
In the lower curve of Fig. 2, values of [ k d - kl - k2/ ( H t ) ] are plotted as a function of the iron(I1) concentrations of the solutions. . I t will be seen that the data satisfy eq. 9 reasonably well. Additional measurements of the iron(I1)-catalyzed dissociation of Fee l 2 +were also made a t constant acidity by using magnesium perchlorate to maintain constant ionic strength. These data are presented in Table I1 and are plotted as a function of the iron(I1) concentrations of the solutions in the upper curve of Fig. 2. Good agreement with eq. 9 is again obtained. The value of ka calculated from the slopes of the lines in Fig. 2 is 12.1 f 1.6 -If-' set.-' a t 25.0" and ionic strength 3.0 .If. , 1.i)
I< E Connick and C . P. Coppel. J A m ( h e m S o c , 81. 6380 (l!i,\(r)
CHLORIDE-CATALYZED IRON(II)-IRON(III) ELECTRON EXCHANGE
Kov. 5 , 1BB1
TABLE I1
- I
I
l
l
/
4593 I
1
k THERATEOF DISSOCIATION OF FeCt2- I N IROS(II) PERCHLORATE-MACN PERCHLORATE-PERCHLORIC ESI~~~ ACID" (Fez'), .M
rnsec.
t,,~,
-
k d , sec. - 1
219 159 185 146 113 95 92
0 025 0 050 0 060 0 10 0 20 0 311 0 40
3.16
-
4 36
3.75 4.76 6.10 7 29 7.53
-
" Mixtures a t 25.0" and ionic strength 3.0 M ; (HClOP) = 1.80 .If.
The Chloride-Catalyzed Iron(I1)-Iron(II1) Exchange. -The dependence of the rate of the iron(I1)-iron(II1) exchange on the chloride concentrations of the solutions is presented in Table 111. It will be seen t h a t the rates TABLE 111 THEEFFECT O F CHLORIDE O S THE RATEO F T H E IROS(I1 )-IRos(III) ELECTRON EXCHANGE' [Fe!IC I x 104,
[Fer 111) ] x 104,
3 6 1 1 1 3 2 1 1 1
00 00 50 50 50 00 00 50 50 c50
1 2 0 0 0 1 1 0 0 0
x
11:2,
k, .tf
.M
sec.
sec. - 1
00 00 50 50 50 00 00
2 04 4 32 4 32 4 32
50 50 50
8 65 10 2
65 33 128 108 90 45 58 77 70 65
26 26 27 32 38 38 39 45 49 53
(CI-)
.ti
.CI
102,
6 50
" .It 25 0" and ionic strength 3 0 .U; IHCIOd) [Mg(CIO4)21 = 0 40 Jf
=
-1
7 3 1
0 5 5 8 0 5 3
180 M;
increase with increasing chloride concentration. The measurements can be interpreted in terms of the rate iaw2,7
R
=
-
+
+
k4(Fe2+)(Fe3+) kj(Fe2+)(FeOH2+)
+ k7(Fe2-)(FeC12+)
k6(Fe2t)(FeC12+)
(12)
from which it follows that
4 0 0
0102
Fe(H,0)83-
+ C1-
and (H?O2FeCl2-
+ C1-
(HLl),FeCIZ'
-
J
K1
(H20),FeC12+ K ,
(14) (15)
0!08
0110
0'12
respectively. Hence k6 and k7 can be calculated from the intercepts and slopes of plots of the left-hand side of eq. 13 vs. chloride, provided K 1and K 4are known. According to Woods, et al.,I6K1 is equal to 8.2 -If-' at 25.0' and ionic strength 3.0 M. The value of K4a t this ionic strength is not known, but we shall assume that the value of K1/K 4 does not vary much in the ionic strength Consequently, we have used range of 1.0 to 3.0 K, = 2.7 J - I a t 25.0' and ionic strength 3.0 L1f.17As is evident from Fig. 3, the data satisfy eq. 13 reasonably well. The values of k6 and k7 calculated from the slope and intercept are 57.6 2 Jf-' set.-' and 159 f 10 M-l set.-', respectively, a t 25.0' and ionic strength 3.0 h". The above value of k6 may be compared with the value of 31 -1T-l set.-' a t 25.(loand ionic strength 0.50 AIf.7 The formulation of the third term in the rate law for the exchange as k6(FeC12+)(Fez-) involves an assumption. Strictly speaking, this term is Klk6(Fe3+)(FeY+) (Cl-), ; . e . , i t defines an exchange path which is first order in iron(II), iron(III), and chloride. However, it is very likely that the reactants in this exchange path are in fact FeC12+and Fez- (as well as FeCl+ and Fe3+), rather than Fez-, Fe", and chloride. For example, recent studies of the chloride-catalyzed chromium(I1)iron(II1) reaction have shown that the rate law for this reaction also contains a term k(Fe3+)(Fe2+) (Cl-) and that the FeC12+-Cr2+ path makes by far the biggest contribution to this term. '8 Comparison of the Dissociation and Exchange Rate Studies.-The rate constant k6 includes all the electronexchange paths involving F e z + and FeC12+. I t is thus equal to the sum of the rate constants for the electron transfer and the chlorine atom transfer paths and is given by
*
kg where k is the over-all second-order rate constant in the presence of chloride ions, and ku is the rate constant in the absence of chloride ions, both a t the same acidity. k6 and k7 are the rate constants for the electron exchange of F e ? + with FeC12+ and FeC12+,respectively, and K 1 and K 4 are the equilibrium constants for the reactions
0106
OiO4
=
(2ka
+ ks)
(16)
where ks is the rate constant for the chlorine atom transfer path FeC12+
+ Fez*
ks
FeC12+ kr
+ Fez+
(17)
(16) M. J. M . Woods, P. K . Gallagher, and E. L. King, I n o r g . C h r m . , 1, 5 5 (1962).
( 1 7 ) E . Kabinowitch a n d W. H. Stockrnayer, J . A m . Chem. .Sot.. 64, :3:{.7 (1942). (18) G Dulz a n d 3 . S u t i n , tbid., 86, 829 (1964)
4594
SHUPACK, BILLIG,CLARK,WILLIAMS, AND GRAY
The coefficient 2 in eq. 16 is necessary because both the forward and reverse reactions of eq. 10 contribute to the observed electron transfer. Since the system is a t equilibrium, the rates of these two reactions are equal. Consequently, the forward reaction of eq. 10 is only responsible for half the observed electron transf e r ~ . ' ~Substitution of k g = 12.1 f 1.6 set.-' and k6 = 57.6 f 2 1W-l set.-' gives kg = 33.4 f 2.6 M-l sec.-l a t 25.0' and ionic strength 3.0 M , X comparison of the magnitudes of these rate constants shows that the electron exchange between Fe2+ and FeC12+ proceeds mainly b y a chlorine atom transfer mechanism (provided, as seems reasonable, the third-order term involving Fez+, Fe3f, and chloride may be neglected). The chlorine atom transfer presumably occurs in a chloride-bridged, inner-sphere activated complex, while the electron transfer may occur in either an outersphere activated complex or in a water-bridged, innersphere activated complex. It is of interest t h a t reaction via the chloride-bridged path is only about three (19) J. Silverman, P h . D . Thesis, Columbia University, S e w York, N. Y., 1951.
[CONTRIBUTION FROM THE
DEPARTMENT OF
Vol. 86
times as rapid as reaction via the outer-sphere or waterbridged paths. This small difference, and indeed the relatively small effect of chloride on the iron(I1)-'iron(111) exchange, contrasts markedly with its effect on the iron( 11)-chromium(II1) and chromium (11)-chromium(II1) reactions. In the latter systems, chloride bridging increases the rates of the reactions by factors of lo4 and more than 2 X lo6, respectively.ld Moreover, the chromium(I1)-catalyzed dissociation of CrC12+is slower by a factor of more than lo4than the Cr2+CrC12+ exchange.20 The difference in the chloride effects cannot be ascribed to any difference in the mechanisms of the reactions since these studies show t h a t i t is very likely t h a t the chloride-catalyzed iron (11)-iron(II1) exchange, like the chloride-catayzed chromium( 11)-iron( 111) and chromium(II)-chromium (111) reactions, proceeds via an inner-sphere activated complex. Acknowledgment.-The authors wish to acknowledge helpful discussions with R. W. Dodson. (20) H. T a u b e and E. L. King, J . A m . Chem. Soc., 76, 4053 (1954).
CHEMISTRY, COLUMBIA UNIVERSITY, N E W Y O R K , N E W YORK
100271
The Electronic Structures of Square-Planar Metal Complexes. V. Spectral Properties of the Maleonitriledithiolate Complexes of Nickel, Palladium, and Platinum BY S.I. S H U P A C KE. , ' ~BILLIG,R. J. H. CLARK,RAYMOND WILLIAMS,'^
AND
HARRYB. GRAY
RECEIVED JUNE 8, 1964 The results of molecular orbital calculations of the square-planar nickel complexes of maleonitriledithiolate are reported. Electronic spectra of the nickel, palladium, and platinum complexes are assigned on the basis of the derived energy levels for the dinegative nickel complex. The observed bands are due to d-d, ligand-tometal and metal-to-ligand charge transfer, and intraligand transitions. The spectral assignments are consistent with the known spectra of square-planar halides, cyanides, and simple derivatives of rnaleonitriledithiolate. The orbital parameter AI is reported for all the complexes, affording a comparison of the x2-y2 --* x y splitting for several complexes containing a nickel-group central metal. Evidence is presented for the assignment of the highest filled orbital a s 4a,(xz-y*) in D 2 h symmetry for the dinegative complexes of the nickel group.
Introduction For several years we have been investigating the syntheses, electronic structures, and reaction mechanisms of square-planar metal complexes. Until very recently, the square-planar geometry was restricted to the diamagnetic, d8 complexes of Ni(II), P d ( I I ) , P t ( I I ) , R h ( I ) , Ir(I), and h u ( I I I ) , except for a few paramagsquare-planar Co(I1) and Cu(I1) comnetic ( S = plexes containing fairly complicated ligands. It is clear that one of the important problems in inorganic chemistry is the preparation and study of square-planar complexes containing as many different transition metals and with as many different ground states as possible. From such research we can hope to help determine the relative energies of the molecular orbitals in planar metal complexes, a problem for which many different solutions have been suggested in the past few years. 2 (1) (a) N I H Postdoctoral Fellow, 1963-1964; (b) N S F Graduate Fellow, 1963-1964. ( 2 ) J . C h a t t , G. A . Gamlen, and L. E. Orgel, J . Chem. Soc., 486 (1938). ( 3 ) G. M a k i . J . Chem. Phys.. !28, 651 (1958); 29, 162 (1958); 2 9 , 1129 (1 9.58) ( 4 ) C . J . Ballhausen and A . D. Liehr. J . A m . Chem. Soc., 71, 538 (1959). ( 5 ) LT. I. B a n , A c f a C h i m . A c a d . Sci. H u n g . , 19, 459 (1959).
From the results of our initial theoretical work, we believe t h a t one way to stabilize the square-planar geometry is to involve the metal d,,, d,,, and pz valence orbitals in an e:tensive n-orbital network spanning the entire complex, thus allowing as much delocalization of charge as possible. Two ligands which have good donor atoms and also a considerably delocalized a-orbital system are the dianions of maleonitriledithiol (MNT2-) and toluene-3,4-dithiol (TDT2-), which are shown below. Whether or not for the reason given above,
( 6 ) S. K i d a , J. F u k i t a , K. Nakamoto, and R. Tsuchida, Bull. Chem. Soc. J a p a n , 31. 79 (19.58). (7) ( a ) R. F. Fenske, D. S. Martin. and K. Ruedenberg, I n o r g . Chem , 1, 441 (1962). ( b ) D. S . hlartin and C . S. Lenhardt, t o be published. (81 J. R. Perumareddi, A . D. Liehr, and A . W. Adamson. J . A m . C'hrm. So,.,86, 249 (1963). (9) J . Ferguson, J . Chem. Phys., 3 4 , 611 (1961). (10) C. K. J@rgensen,"Orbitals in Atoms and Molecules," Academic Press, Piew York, X. Y.. 1962, p. 114, and references therein.
